DEC 9 Track Tapes (was Re: Applikon Workstation?)

[Jay Jaeger]

On 8/29/2015 10:57 AM, Al Kossow wrote:
> On 8/28/15 9:25 AM, Jay Jaeger wrote:
> 
>> Many/most 9 Track tapes (those from the early to mid-eighties until 1995
>> or so - what matters is the date of manufacture, not when they were
>> written) have to be "baked" before reading, owing to "sticky shed
>> syndrome".  My experience with tapes earlier than that is that I can
>> read them without baking them first.
>>
> 
> Mid-80's are the worst, especially Memorex, and BASF, which everyone used
> because they were cheap.
> Whatever HP bought for their distribution tapes (probably Graham
> Magnetics) is very good.
> IBM tape is good too.
> DEC used crap tape for their distributions.
> 
> Sadly, AT&T and Mt XINU ALSO used especially bad tape in
> the late 80's, so many of the Unix distributions I've been dealing with
> from
> them lately are so sticky even after baking that they are impossible to
> deal
> with.
> 

You can add Burroughs SP 3200 and SP 5000 and SP 9000 to your "bad boy"
list.

I had good luck with a 3M 777 6250 BPI Security Computer Tape.

Like you, I have not had any issues with IBM tape.

One Graham Verituf Certified Zero Defect at 3200 FCI / 1600 BPI that I
processed did leave some residue, even after baking.

If a tape is at 1600 BPI, I can sometimes get through it even if it is
sticky after baking, so long as there are no retries.  In fact, I have
taken to setting retries to the smallest possible number (I think it is
2) on the drive during my first attempt at reading a tape so that if it
does stick, I minimize the damage.  A few times I have had luck with
doing that, having a couple of read errors, then cleaning the drive  and
reading the tape again with the normal number of retries (9).

At 6250 BPI, the drive won't stream continuously, so if one of those is
sticky after baking, it is usually a challenge to get a good copy.

JRJ